1. Introduction
2. Team Structure3. Process Flow
3.1. Module Learning & Experience Design
3.2. Individual Flow Design
3.3. Narrative Design & Skills Mapping
3.4. New Assets Creation
3.5. Annex: World Building - Learning & Experience Design
3.6. Annex: World Building - Story Design
3.7. Process Flow Conclusion4. Caveats
4. Caveats5. LLM Prompt Engineering Techniques
5.1. Use The Latest Model
5.2. Zero-shot Prompting
5.3. Few-shot Prompting
5.4. Chain-Of-Thought Prompting
5.5. Structuring Prompts
5.6. Describing Prompts
5.7. Editing Prompts
5.8. Extending Responses
5.9. Multiple Users Collaborating6. Text-to-Image Prompting Engineering Techniques
6.1. General Techniques
6.2. Photography
6.3. Architecture
6.4. Various Aesthetic Styles
6.5. Product & Material3.4. New Assets Creation
This process of creating new assets may not be applicable to you. Please check with the Talespin leads if you would like to create new assets. The Narrative Designer will develop the new assets.
3.4.1. Developing Character
This step is the same as in process ‘2. Individual Flow Design’ step ‘2.1. Character & Environment Identification’.
However, this creation process is not restricted to the existing characters available in CPD. We can design a new virtual character with a new physical appearance, gender, and race.
3.4.2. Developing Environment
Design a suitable environment in which a learning module about [insert topic] is conducted, based on the following information provided.
Module description: ###
[Insert module description]
###
Character profiles: ###
[Insert character profiles, and their relationship with each other and the learner]###
Adhere to these guidelines: ###
1/ The environment should have physical elements that facilitate the discussion of these topics.
2/ This environment should embody elements such as [insert elements, if any], and values such as [insert values, if any].
3/ Provide detailed and creative descriptions of the environment. Feel free to be as imaginative and descriptive as possible. The more detailed and imaginative your description, the better it is.
###
Other than the straightforward approach above, there are 2 methods that can be used to obtain inspiration for the desired environment.
- Scenarios that describe the topic
- Analogies that explain the topic
Provide a list of scenarios that best describe the [insert the topic].
Provide a list of analogies to explain the [insert the topic].
Knowing about scenarios and analogies related to the topic will provide us with insights into suitable settings. For instance, in the case of conflict resolution:
- An example of a scenario would be a customer who is unhappy with a product or service and is expressing their complaints to the customer service representative. With this in mind, we could create a few settings: the company office, warehouse, service center, and the customer's home.
- An analogy that illustrates conflict resolution could be compared to caring for a garden. Just like a garden needs to be tended to in order to flourish, conflict resolution requires addressing and resolving conflicts in order to promote healthy relationships and growth. To represent this analogy, we could have settings such as the company office and the company's garden.
3.4.3. Character Physical Appearance
This step could be skipped if we prefer to create a character directly with the text-to-image tools without the aid of ChatGPT.
Design and describe [Insert character’s name]’s physical appearance.
Character’s profile: ###
[Insert character’s profile]
###
Elaborate on the description of [Insert character’s name]’s physical appearance and abilities, including skin color, hair color, and style, eye size and color, skin complexion, nose structure, ears, mouth, jawline, and more.
3.4.4. Environment Physical Appearance
This step could be skipped if we prefer to create a character directly with the text-to-image tools without the aid of ChatGPT.
Provide detailed and creative descriptions of the physical appearance and elements of [insert description of the environment].
Feel free to be as imaginative and descriptive as possible. The more detailed and imaginative your description, the better is.
3.4.5. Character Artwork
We could keep it open-ended with the text-to-image tools with simple prompts such as “A 60 years old British lawyer” and leave it to the tools to do the job.
Otherwise, we could use the keywords provided by ChatGPT in an earlier step for the artwork:
Portrait, [Insert description of the character’s physical appearance]
E.g.
Portrait, 60 years old British woman, medium-toned and olive-colored skin with a smooth complexion, shoulder-length wavy chestnut brown hair, almond-shaped and hazel color eyes, well-defined eyebrows, the nose is slightly aquiline with a straight bridge, strong jawline, and well-defined cheekbones
Output:
Shortlist your preferred image.
Output:
3.4.6. Environment Artwork
We could keep it open-ended with the text-to-image tools with simple prompts such as “A modern courtroom” and leave it to the tools to do the job.
Otherwise, we could use the keywords provided by ChatGPT in an earlier step for the artwork:
Modern courtroom with a raised platform. On the highly raised platform, there’s a judge’s bench which is a massive, ornate piece of furniture that is carved from fine mahogany and accented with gold leaf. On the judge's bench is a large gavel. wide-angle view, cinematic lighting
Output:
Refine further.
Output:
3.4.7. Merging Artwork
Next, we merge the characters and environments artwork. There are 3 methods below where they are ranked in the order of their effectiveness.
Method 1: Upload Both Characters And Environment Images
[Insert URL of character image], [Insert URL of environment image], Combine both images, including the woman in the courtroom, hyper-realistic, cinematic lighting, HD
Output:
Method 2: Only Upload The Environment Image
In this method, we only upload the environment image while we enter the prompt of the character.
[Insert URL of environment image], use the courtroom as the background, include a 60 years old British woman, with medium-toned and olive-colored skin with a smooth complexion, shoulder-length wavy chestnut brown hair, almond-shaped and hazel color eyes, well-defined eyebrows, the nose is slightly aquiline with a straight bridge, strong jawline, and well-defined cheekbones
Output:
Method 3: Only Upload The Character Image
In this method, we only upload the character image while we enter the prompt of the environment.
[Insert URL of character image], including the women in a modern courtroom.
Output:
3.4.8. Documenting Artwork
With the generated artwork, we would compile them into a simple document, which will be used to communicate with the Art team on what is expected of the virtual assets.
Download the Concept Deck template in the ‘Resources’ section.
3.4.9. Learning Designer, Development, SME & ART Review
The Producer will arrange for the Learning Designer, Developmental Editing Lead, relevant Subject Matter Experts (SMEs), and ART Team to review the material simultaneously.
These reviewers will have a 48-hour window to offer their feedback. Should they fail to do so, the Producer reserves the right to move forward without their input.
After receiving the reviewers' feedback, the Producer will instruct the Narrative Designer to implement the required modifications.
3.4.10. Stakeholder Approval
Once the Narrative Designer has made the necessary changes, the Producer will arrange for the final review by the final stakeholder who is usually one of the Learning Experience Leads.
On this page:
- 3.4. New Assets Creation
- 3.4.1. Developing Character
- 3.4.2. Developing Environment
- 3.4.3. Character Physical Appearance
- 3.4.4. Environment Physical Appearance
- 3.4.5. Character Artwork
- 3.4.6. Environment Artwork
- 3.4.7. Merging Artwork
- 3.4.8. Documenting Artwork
- 3.4.9. Learning Designer, Development, SME & ART Review
- 3.4.10. Stakeholder Approval